Viewpoint: Artificial intelligence poised to play greater role in science

28 Apr 2022 | Viewpoint

Global survey shows researchers in increasing numbers are looking to AI to support and speed up their work

Adrian Mulligan is Research Director for Customer Insights at the science publisher Elsevier

COVID-19 changed the way we all work, live and socialise – with technology and communication tools more important than ever. The world of scientific research was no exception, with the use of technology and specifically artificial intelligence (AI) increasing. AI is now increasingly relied upon to speed up research and generate new insights across all of science.

The positive endorsement of AI was noted in the second iteration of Elsevier’s global research project – Research Futures – which aims to gather the views and opinions of researchers across the world to help us as science publishers better understand the challenges and opportunities they face.

Forty seven percent of researchers believe that a long-lasting impact of the pandemic will be a greater dependency on technology and AI in their work, underlining the importance of AI for the future of research.

The study shows the number of researchers using AI “extensively” has increased from 12% in 2020 to 16% in 2021. In materials science, which covers the structure and properties of materials, the discovery of new materials and how they are made, 18% of researchers are now likely to be extensive AI users, up from zero a year ago. In chemistry, the number has grown from 2% to 19% and in mathematics from 4% to 13%. Unsurprisingly, 64% of computer scientists say they are heavy users of AI. 

Most often researchers who use AI do so to analyse research results (66%) or to spot defects or issues with data (49%), while a minority are using it to help generate new hypotheses (17%).

As we note in our Research Futures Report 2.0, AI has been crucial to healthcare throughout the pandemic. We have seen hospitals use it to help predict which patients would be most severely affected by COVID-19, as well as manage their resources.

Attitudes toward the use of AI in peer review have also changed. Around one in five researchers (21%) agree they would read papers that rely on AI for peer review instead of humans, a 5% increase on 2020. Looking at the results by age, those aged 36 and under have increased their willingness to read such articles the most, compared to a year ago (21%, up from 14% prior year). But while attitudes are changing, most researchers continue to be reticent about AI in peer review, with 58% saying they are unwilling to read such articles.

It’s clear that the place of AI in research is evolving and it is gradually becoming a crucial and trustworthy tool. However, not all reservations surrounding AI have been answered by the accelerated reliance on it during the pandemic.

Nonetheless, the technological strides made, especially in the fields of material science, medicine and chemistry, show the crucial role AI will play in the future.  

The Elsevier Research Futures Report 2.0 is free to download here. It builds on the first Research Futures Report (2019) which considered what the world of research might look like in 10 years’ time. The new data highlights mounting pressure across publishing and funding, while highlighting new opportunities in new funding sources, technology, and collaboration.

Adrian Mulligan is Research Director for Customer Insights at the science publisher Elsevier

Never miss an update from Science|Business:   Newsletter sign-up